Comment by Gpetrium | Hacker News Reader

Comment by Gpetrium | Hacker News Reader

Parent Full thread

@kartan, there is a lot of valid reasons to give an incentive such as a tax break to a large company, I will list a few:

1. Brand name corporation in your community, leading to new businesses propping up

2. Increase land-value in the area, leading to higher city taxes

3. Income, corporate and other taxes associated with the company's project & continuous operation

4. Long-term certainty. Most companies wont build a data center only to scrap the project in 2 years.

5. Economic multiplier. If we assume that each job is tied to $100,000 in income, taxes, benefits, etc and there are 2,300 jobs for 24months, we are looking at $345M-460M in 2 years. Run an economic multiplier and you may be reaching the billion mark

6. Plenty of city,state,federal projects run at a negative or neutral economic value. Having a business such as Google offering to bring business to your hometown will certainly make you listen to them and consider their offer.

It becomes a matter of cost/benefit analysis and ensuring that the community accepting this transaction understands and has the option to complain about it.

I agree that if everyone plays this game, we may reach a point where governments are fully subsidizing every business (there are already plenty of businesses subsidized by governments). It is a matter of finding the right balance.

1. There are already dozens of datacenters in Minnesota. They aren't prompting other businesses to pop up because they are self-contained and self-serving. Furthermore there is no "brand recognition" they are giant unmarked warehouses that could be a datacenter or could just be another nameless storage facility.

2. Why would a datacenter increase land value? This isn't a corporate headquarters, it's a datacenter.

3. Taxes on what? That's what this package is, tax exemtpion.

4. Again, the datacenter employs "20 people" which is really like 4 FTEs and a bunch of cleaning staff. And likely part-time Security guards.

5. There won't be 2,000 jobs to build a datacenter, that's wildly exaggerated. And there's no way in hell that the average construction worker is making even half of the number you just stated.

6. If they were bringing actual jobs, and not a borderline-unmanned datacenter, sure. This isn't that.

I will bite your response with the hopes you try to keep an open mind to the response you get instead of a preconceived notion of how things work.

1. Data centers are a symbol to a much larger economic schema that impacts quality of service to the area, economic prosperity, among other areas. Minnesota can literally come to prospect business and say "Look, Google, X, Y and Z companies gave us a seal of approval that our state is tech driven, come join them in the future of technology in America" You would be surprised how 1 transaction with a big brand can help future business opportunities.

2. One data center alone does not increase land value, the combination of all the variables I list ++++ do.

3. Every employee that is paid is taxed, every equipment they purchase is taxed, every goods or services they need is taxed, land is taxed. Taxation is not just what income tax and corporate tax.

4. You shouldn't be comparing a data center to a headquarter, they are two distinct parts of a business and the economy. You will have 20 people + the contractors needed from time to time + folks that need to fly to the data center 70% of their time + other marketing, sales jobs to sell the data center services to the state economy. Is it on the same degree as a headquarter? Never, nor should it be seeing as such

5. You could divide the number of jobs by 30 and you could still break-even as a city/state without accounting for the other benefits I list (and the ones I didn't). You are also forgetting that there are other costs associated with the job that are not construction work and even within construction work, there are specialists that make more money. Always remember, just because your paycheck says you make $15 an hour, it doesn't mean that your cost is a lot higher.

6. Refer to the other points and try to think of the ones I didn't list (improved disaster readiness in the region, energy consumption cost and taxes/jobs associated with it, promises to set-up green initiatives, etc)

I hope this has helped you see a different side of the equation.

This is a very good discussion, back and forth. I think it is worth asking who should be capturing the economic benefit here. A tax break for google in this case is allowing google to capture the economic benefit it creates for the public. Not giving it a tax break allows the public to capture that economic output. While I very much agree that this project will add value to the economy, if google is given a $15m tax incentive to create this property, it means that the public doesn't get it.

Those justifications you lay out are real, but if you pay google for them in the form of a tax break it all balances out. It doesn't create value from nothing.

Maybe, it could also mean that Google go to another state who will offer a tax incentive, and the state loses out.

Or, as AOC would say: saves $15m.

"the state loses out" assumes that the public makes money off the deal. If not, just spend the $15m on programs that actually benefit them.

> Those justifications you lay out are real, but if you pay google for them in the form of a tax break it all balances out. It doesn't create value from nothing.

Only under the assumption that $15M is the exact maximum amount that the city would pay for the listed benefits (i.e. the exact break even number). Probably it is worth more than that to the city (or else they probably wouldn't do the deal), so the reality is that Google is capturing some of the surplus value, and the city is capturing some.

>Google is promising to bring 50 full-time positions to the area. The company also projects that it will create 2,300 jobs during the 18–24 month construction window—a far shorter period than the 20 years of tax relief they're requesting. After construction wraps, with 18 remaining years of tax abatement, there will be a $270,000 tax shortfall for every full-time position created.

I read that paragraph from the article to read that after the 20 year term google is looking for this tax relief, the town will be net negative 270K for each of the 50 employees Google creates. If that is not what that paragraph means, then I agree with you, but, my interpretation is that for the state to break even, google has to generate an additional 270K of economic return to the state per each person at the google data center. I find that number to be unlikely.

I read that too and don't understand it at all. Do you? By "shortfall", do they simply mean "if we charged full property taxes to Google, the revenue we'd get would be much more than the taxes we'd get from the 50 FTEs at the data center"? "Shortfall" suggests money the city and county would essentially be "spending" to have the data center, but that's not what's happening, right?

https://www.taxnotes.com/editors-pick/google-seeks-20-years-...

that seems to be a much better article, but still doesn't elaborate on that "shortfall". The Taxnotes article makes this seem like a much better deal for the down, and a much more reasonable ask from google. They are effectively asking that they be granted a reduced increase in taxes on the proposed vast increase in property value from the value of their development"

>The county assessor valued the land at $1.76 million, and the city estimated that the property's market value upon completion would be $41.6 million. Local real estate taxes on the property are around $16,476 and are estimated to jump to $328,726 after the facility is built. Jet Stream requested property tax breaks for 20 years for the data center, but not school district tax abatements.

What a great find. Thank you!

I'm sorry, I can't comment on the specifics of the deal. I'm only trying to point out that the exact total amount of surplus is not necessarily equal to the amount of tax abatement Google is requesting (it could be more or less).

it's 15MM over 20 years, right? $750k/year? Against the economic activity from building a huge data center, that seems like a drop in the bucket?

You presume they don't have other options. If they don't give any incentive, then they just go somewhere that does offer incentives. Wisconsin or North Dakota or whoever else will take them.

Data centre placement is driven by a heck of a lot more than tax incentives.

If they're looking to build a data centre in one place its because it fits their huge list of criteria. Tax breaks are just icing on the cake. They want a data centre here regardless.

I'm really curious what you think the other criteria are, because I can tell you that Becker, MN, is around 90 minutes from the Minneapolis-St. Paul metro area and solidly in the middle of farm country.

I can only surmise that their list of criteria was limited to "minimize dollars per square foot in rent per month" and "vaguely somewhat near a metro area."

Power is the biggest expense of a data center. So either access to cheap power, or access to cheap cooling to offset power are probably the top criteria.

I'm sure there's others, latency to customers and real estate cost like you mentioned. Connectivity, construction costs, political stability.

Can you list any data showing datacenters increase land value? Or have boosted an economy? Or are these theoretical benefits?

I just found this pdf from the US Chamber of Commerce that can shed some light on a few questions you may have (HTTPS+Valid Secure Certificate):

Data Centers: Jobs and Opportunities in Communities Nationwide https://www.uschamber.com/sites/default/files/ctec_datacente...

It covers: Initial capital and operational expenditure, economic impact, spillover benefits, etc

In addition to Gpetrium's comment, there is this report by Oxford Economics that has been linked to elsewhere in the comments: https://static.googleusercontent.com/media/www.google.com/en...

All of the above assumes the people Google is hiring are unemployed which is simply not going to be the case.

The current unemployment rate in Minnesota is 2.8%. The current unemployment rate among tech workers is less than 1%. This will not generate anywhere near the pie-in-the-sky numbers presented in any findings. At best it will ever-so-slightly increase the tax base by giving a small raise to a tiny handful of individuals. The workers building the actual datacenter will almost assuredly be in-state construction workers whose salaries will not change as a result of the project, this is just one more project to add to the list.

Furthermore, once again, Minnesota already has a glut of datacenters both public and private. none of the "other benefits" even apply. They aren't putting this up in the middle of nowhere Montana. While it is admittedly presumptuous of me, I'm going to assume you have absolutely no idea what the tech industry in Minnesota looks like.

> 5. There won't be 2,000 jobs to build a datacenter, that's wildly exaggerated. And there's no way in hell that the average construction worker is making even half of the number you just stated.

Most union tradespeople make at least $35/hr in MN, a handful make over $40. My company pays the union about $75/hr for a journeyman electrician, which is their full rate with benefits included. Most (or all?) the labor on this project will be union labor, a GC isn't going to take a chance on a non-union sub with 600 million on the line.

> 2. Why would a datacenter increase land value? This isn't a corporate headquarters, it's a datacenter.

Data centers require backbones to provide their services. Becker sits outside the range of most ISP's, in addition to no fiber connection; Monticello being the closest with fiber. I suspect there will be a substantial contract for laying fiber here, thus further increasing value of the land.

Centurylink is already passing through there. They already have 1gb internet for 65.00/month there. That's cheaper than what I pay here in Austin.

They dont have GB internet there, I have family there. =P

>There won't be 2,000 jobs to build a datacenter, that's wildly exaggerated

There could be, they probably didn't pull that number completely out of nowhere. But there certainly won't be 2000 jobs for the duration of the construction. or 2000 new jobs.

I'm guessing it just creative counting, like if a crew of four people comes out to plant a couple trees by the front door, works for 4 hours total, and has no other involvement in the project, that's 4 jobs.

Reply on news.ycombinator.com